The invention relates to gas turbine engines, particularly the internal cooling of turbine blades, and to improved method and apparatus for delivery of cooling air to the blades.
In essence, a gas turbine engine comprises a compression stage in which air is pressurized. Pressurized air is sent to a combustion chamber where it is mixed with fuel and the mixture is ignited. The combustion gases produced by the ignition of the air/fuel mixture is a hot, rapidly expanding gaseous volume which is directed from the combustion chamber to a turbine wheel to drive the latter.
In order to achieve efficiency available with higher temperature gas turbine engine operation, the turbine rotor blades are effectively and efficiently cooled. The prior art has provided rotor blades with internal passages. Cooling fluid is injected into these internal cooling passages or channels, flows through the interior of the blade and exits therefrom to provide additional surface cooling of the blades. This arrangement is disclosed in greater detail in U.S. Pat. No. 4,270,883, assigned to the same assignee herein, issued Jan. 2, 1981; made a part hereof by reference.
When considering the addition of apparatus to cool the rotor blades of a turbine engine, the operational speed at which parts of the engine move must be given due consideration. For example, the rotor blades move in a circular path as the rotor rotates. Within as little as one second, the tip of each rotor blade may travel 2,000 feet or more. Under such dynamic conditions it is of utmost importance that the structural integrity of the rotor/rotor blade assembly be maintained.
In general, the coolant fluid employed in a gas turbine engine will be air derived from the compression stage. The more air that is drawn off for cooling purposes, the less air available for use in the combustion chamber. The most efficient cooling system will draw the least amount of air from the compressor. Once air has been drawn from the compressor for use as a coolant, it is an important function of the cooling system that losses, both fluid and mechanical, be minimized.
Cooling fluid flow losses are generally attributable to insertion losses and pumping losses. Pumping losses result from the energy taken from the rotor as it works to move coolant fluid from the radius at which it is injected on or into the rotor outward to the radius of the rotor blades, at the periphery of the rotor.
Insertion losses are made up of seal losses, frictional losses, and swirl losses. In directing coolant fluid from one place to another, the fluid may move from regions having stationary elements to those having rapidly rotating elements. Because of the high speed of rotation of these latter elements, the seals employed between stationary and moving parts generally comprise labyrinth structures which impede the flow of coolant fluid through them by providing a high impedance, tortuous fluid-flow-path. However, the structure is basically a leaky one and becomes more so as the pressure of the fluid impinging on the seal increases.
Seal losses are minimized by minimizing the static pressure of the fluid impinging on the seals. Reduction of seal losses, in turn, reduces the amount of air that must be supplied by the compressor stage. For a given compressor stage, there is more air available for combustion purposes when seal losses are minimized.
Frictional losses result from the interaction of the rotational elements of the engine with the coolant fluid. Frictional losses reduce the efficiency of cooling by raising the temperature of both the coolant and the moving parts, and by decreasing the coolant's pressure. Thus, when frictional losses are significant, a higher initial coolant pressure is required. This higher pressure increases the burden on the seals and an increased seal loss derives.
Swirl losses are caused when the rotor, rotor blades, or other rotating parts of the turbine engine have to impart energy to the coolant to accelerate the coolant fluid such that the coolant itself acquires a rotational velocity or swirl equal to that of the rotor or other rotating part. This places a load on the rotating parts, raises the temperature of the coolant fluid and reduces the shaft energy available from the turbine engine.
An optimum cooling system will minimize insertion losses (seal, frictional and swirl) and pumping losses by minimizing the work done in moving coolant fluid into the rotor blades, all the while maintaining structural integrity of the rotor. The art in this area has concentrated for the most part on a single approach to more efficiently supply cooling air to turbine blades, namely, by imparting some degree of swirl to the cooling air before it is supplied to the turbine rotor, thereby minimizing some portion of the insertion losses. This technique will reduce swirl loss, for, if it is performed effectively, the cooling air is brought to a tangential velocity equaling the tangential velocity of the turbine rotor at the point at which the cooling air is supplied to the turbine rotor.
An early reference utilized this approach is U.S. Pat. No. 2,910,268 to Davies et al., which is an apparatus for tapping air from a compressor section of a turbine engine and providing it to the interior portion of the shaft of a turbine rotor. Succeeding references have further improved the techniques of preswirling the cooling air so as to reduce some components of insertion losses including swirl loss. Such references include U.S. Pat. Nos. 2,988,325; 3,602,605; and 3,936,215. By preswirling the cooling air, insertion losses are reduced somewhat. Additionally, the preswirling effectively presents nozzles through which the cooling flow expands to reach rotor speed at a lower pressure and temperature.
These devices all possess significant problems in delivering the cooling air to the turbine blades in that they require a primary design choice to be made. If cooling air is supplied at high pressure to the turbine rotor, there is a substantial leakage problem resulting in the loss of a significant percentage of the cooling air and resulting in reduced efficiency in the cooling operation. The other alternative involves supplying cooling air at a somewhat lower pressure and utilizing a pumping vane to move the air from the interior of the turbine rotor outward to the turbine blade. This technique necessarily involves performing a substantial amount of work on the cooling air, decreasing the efficiency of the cooling operation and causing drag on the turbine wheel as well as increasing the temperature of the cooling air supplied to the turbine blades. An example of such a pumping blade is shown in U.S. Pat. No. 3,602,605 to Lee et al.
Very high speed rotating turbomachinery has yet further trade-offs in structural design considerations for delivering cooling flow to rotating, internally cooled turbine blading. Specifically, the optimal radius for injecting cooling flow into the turbine rotor oftentimes lies between the central bore of the rotor and its outer periphery where the turbine blading is located. Because of the high rotational speed of the rotor, however, in smaller gas turbine machinery the necessarily rotating support structure for defining an entry for the cooling flow at the optimal radius must be quite heavy and bulky in order to withstand the high centrifugal loads thereon. This results in added weight and dramatically increased mechanical complexity for the smaller gas turbine engine.
Additionally, it has been found that such rotating support structure must be axially offset somewhat from the rotating turbine in order to provide the required strength to these rotating elements. As a result, the cooling flow is then necessarily injected into a relatively large, open cavity wherein various aerodynamic losses occur to the cooling flow before it can find its way to the passages ultimately leading to the internal turbine blade cooling passages.